Blind Channel Estimation and Data Detection With Unknown Modulation and Coding Scheme

Blind signal processing techniques of channel and noise power estimation, modulation and channel coding scheme (MCS) recognition, and data detection have played crucial roles in wireless transmission scenarios, where the receiver cannot obtain a priori information of the incoming signals. Each of these blind signal processing tasks has been extensively studied in the literature. Few works studied the combination of partial tasks jointly. However, to the best of our knowledge, an overall problem that involves all the aforementioned tasks has not been investigated previously. Simply cascading the solution of each individual task is apparently not a suitable approach for the overall problem. To address the above issues, we design a joint blind receiver that jointly improves the performance of parameter estimation, MCS recognition, and data detection. Thus, we refer to it as the blind estimation, recognition, and detection (BERD) receiver. The multipath fading channel, noise power, and MCS of the transmitter are all unknown to the receiver side; the required side information is only the pool of MCS candidates and perfect synchronization. In this BERD receiver, we first propose an expectation-maximization-based channel and noise power estimator block, which solves the non-convex maximum likelihood estimation problem in a tractable way. Then, a soft-information detector and regenerator block is designed to detect the transmitted information bits and regenerate more reliable symbols. The BERD receiver iterates between these two blocks, and consequently, the accuracy of channel estimation is improved and in turn helps the data detection. Finally, a multistage likelihood-based fusion and decision block is proposed to make the final decision on the adopted MCS, the information bits, and the unknown channel information. Numerical results are provided to show the superior performance of the proposed BERD receiver compared to the existing schemes, in terms of data detection, MCS recognition, and channel estimation.